March 5, 1959
1249
A4LKYLCYCLOIIEXYL RROMIDES WIT11 'hIOPHENOL.'.TE
infrared spectrum was identical with that of Eastman Kodak c o . spectral grade isooctane, nzKD1.3380. 2,2,3,3-Tetramethylbutane which solidified in the distilling head during the distillation of the reaction mixture, was recrystallized once from methanol, m.p. 100-101", reported34 m.p. 100.6'. The infrared spectrum was idenreported by the tical with that of 2,2,3,3-tetramethylbutane American Petroleum Institute Project 44 (Spectrum 444). 2,4,4-Trimethyl-l-penteneoccurred in a mixture containing the other Cs-hydrocarbon as well as ethanol. I t could not be separated from the 2,2,3,3-tetramethylbutane,even by gas chromatographic techniques. Therefore, 11 nil. of the distillate in which the 2,4,4-trimethyl-l-pentene was present was diluted with 30 ml. of ether, washed 7 times with 25-m1. portions of water (to remove the ethanol), and twice with saturated salt solution. The ether layer was dried over sodium sulfate and the ether then slowly evaporated. To the residue, in 25 ml. of glacial acetic acid, was added 1.06 g. of benzyl mercaptan (0.0086 mole), 10 mg. of hydroquinone and finally 50 drops of 18 M sulfuric acid. After standing 16 hours a t room temperature, the reaction mixture was diluted with an equal volume of water and extracted with two 35-ml. portions of CClr. After drying the extract over sodium sulfate, the CCL, as well as anything else volatile at 100" (10 m a . ) , was removed by distillation. The oily residue, which was void of any mercaptan odor, was oxidized with 10 ml. of peracetic acid. The addition of water to this oxidation mixture resulted in a white precipitate, 0.425 g., m.p. 120.3-121.0", after 3 recrystallizations from 30% ethanol. This was identified as benzyl 1,1,3,3-tetraaethylbutylsulfone, since it did not depress the melting point of known material (see above). (34) S. W. Ferris, "Handbook of Hydrocarbons," Academic Press. Inc., Xew York, N. Y . , 1955, pp, 25-256.
Thermal Reaction of DTBP with Triethyl PhosphiteDTBP (31.6 g., 0.216 mole) and 121 g. of triethyl phosphite (0.73 mole) were heated below a reflux condenser for 42 hours a t 120". Nitrogen was bubbled through the reaction mixture continuously, and thegaseous stream passed through a trap a t -80". Gas chromatographic examination of the contents of the trap revealed a mixture of low (C,) and high (C,) boiling hydrocarbons, very similar to the mixture obtained in the photolysis of DTBP in triethyl phosphite, with the exception that apparently no 2,4,4-trimethyl-l-pentene formed, but instead a peak corresponding to authentic 2,2,4-trimethyl-2-pentene. A white crystalline solid, 0.53 g., whose infrared spectrum was identical with that of 2,2,3,3-tetramethylbutane, except for medium sized peaks a t 9.7 and 10.9 fi (where triethyl phosphite has very strong bands) was found at the inlet to the trap. After 2 sublimations (atmospheric pressure, 50") a product, m.p. 93-97', was obtained. The infrared spectrum was again identical with that of 2,2,3,3-tetramethylbutane;however, the peaks a t 9.7 and 10.9 fi were still present, although greatly reduced in relative size. Fractional distillation of the reaction mixture revealed 86.4 g. of triethyl phosphate (11Oyg based on oxygen available from the peroxide), b.p. 126 ~ 16 g. of material boiling above the (40 mm.), n Z 51.4048; triethyl phosphate was found, but not identified. Reaction of &Butyl Hydroperoxide with Triethyl Phosphite.-&Butyl hydroperoxide was added dropwise to 9.7 g. of triethyl phosphite a t 0". Reaction was instantaneous and exothermic and addition was continued until no further heat was produced. Gas chromatographic analysis of the reaction mixture revealed that only two products formed, tbutyl alcohol and triethyl phosphate. When the reaction was repeated in the presence of 0.100 g. of trinitrobenzene ( a pink solution with triethyl phosphite), an exactly analogous instantaneous and exothermic reaction occurred. NEWYORK27, AT.Y.
[CONTRIBUTION FROM THE CONVERSE hfEMORIAL LABORATORY O F HARVARD UNIVERSITY AND O F THE UNIVERSITY OF NOTREDAME]
THE
CHEMICAL LABORATORIES
Conformational Analysis. VII. Reaction of Alkylcyclohexyl Bromides with Thiophenolate. The Conformational Equilibrium Constant of Bromine1 BY ERNESTL. ELI EL^ AND RALPHG. HABER RECEIVED AUGUST26,1958 The bimolecular substitution and elimination rates of cyclohexyl bromide ( k ) , cis-4-t-butylcyclohexyl bromide ( k , ) and trans-4-t-butylcyclohexylbromide ( k , ) with thiophenolate have been measured. The over-all rate for the (axial) cis isomer is about 61 times as large as that for the (equatorial) trans isomer, the ratio for the corresponding tosylates (ref. 8) being 36, The equilibrium constant: axial cyclohexyl bromide F? equatorial cyclohexyl bromide, calculated by the previously presented (ref. 3 ) equation K = ( k , - k ) / ( k - k , ) is 3.4, corresponding to a difference of 0.73 kcal./mole between equatorial and axial bromine. This difference, which is supported by data in the literature, is considerably smaller than the accepted difference of 1.6-1.8 kcal./mole between equatorial and axial methyl. It is suggested that the difference is due largely to London forces which, at the inter-atomic distance between axial positions in cyclohexane, are of similar order of magnitude as the repulsive van der Waals forces. Contrary to'an earlier report (ref. 8), elimination occurs in the reaction of trans-4-tbutylcyclohexyl tosylate as well as trans-4-t-butylcyclohexyl bromide with thiophenolate. This observation supports the "merged mechanism" of Winstein, Darwish and Holness (ref. 36) for the thiophenolate displacement.
I n a previous paper3s4 we have proposed the relationship K = (ka - k ) / ( k a - k e ) (i) where K is the conformational equilibrium constant for the atom or group X (Fig. l),ka and k e are specific rates of a suitable reaction5 for the axial and equatorial (1) (a) Presented, in part, before the Division of Organic Chemistry, Sdn Francisco Meeting, Am. Chem. Soc., April 15, 1958; (b) Paper VI. 1.O w . Chem., in press. (2) National Science Foundation Senior Postdoctoral Fellow 19581959. (3) E. L. Eliel and C. A. Lukach, THISJOURNAL, 79, 5986 (1957). (4) S. Winstein and N . J. Holness, i b i d . , 77, 5562 (1955), have earlier proposed an equivalent relationship k N&, N.k. where N e and N . are the mole fractions of substrate in the equatorial and axial conformation. respectively. ( 5 ) The nature of the reaction chosen is immaterial.
-
+
isomer, respectively, and k is the experimental rate constant of the naturally occurring equilibrium mixture of the two conformational isomers for the reaction in question. The constants k, and k, are measured experimentally by studying the reaction in question with 4-t-butylsubstituted compounds.* Because of the large bulk of the 4-tbutyl group,4 this group will tend to occupy exclusively the equatorial position. Therefore, in cis-4-t-butylcyclohexyl-X,X will occupy the axial positiona and the compound will react a t the specific rate ka, whereas in trans-4-t-butylcyclohexyl-X, X will occupy the equatorial position (6) Provided X is substantially smaller than I-butyl.
1250
ERNEST L. ELIELAND RALPHG. HABER
lka
lke -
Products
Products
Fig. 1.
'and the compound will react a t the specific rate ke.7
Vol. 81
the over-all rate was followed acidimetrically and the substitution rate alone was followed iodimetrically. However, since previous work had indicated that in slow runs of the type encountered in the present study (extending over several days), extensive oxidation of thiophenolate interfered with the iodimetric titrations, the present work was carried out in ampoules sealed under nitrogen. This gave steady and reproducible ratios of k s / k ~(ks = substitution rate, kT = total rate, k~ = k~ - Ks = elimination rate). Average rate constants for our runs are given in Table 1.'' Included in Table I are data earlier obtained for n-butyl bromide and sec-butyl bromide in stoppered flasks.'* The individual runs are shown in Table I11 (Experimental) and a typical run is detailed in Table IV.
Previously, equation i has been used to calculate the conformational equilibrium constant for hydroxyl (from oxidation4 or acetylation3 rates) and for the tosylate group (from solvolysis rates4 or TABLE I bimolecular reaction rates with thiophenolates). The present paper reports the application of the AkT.ERAGBD RATECONSTANTSn FOR ALKYL BROMIDESK I T H same treatment to a calculation of the conformaTHIOPHESOL.4TE IS 87% ETHASOL AT 25.1' tional equilibrium constant for bromine. This Alkyl group k r X 105 lz,,'irr k s X 105 kE X lo5 seemed of interest since a previous estimateg 12 -Butyl 1750 1'' 1750 0 G8.2 c u . 1' ca. 6 8 . 2 ca. 0 of the energy difference E B between ~ equatorial and sec-Butyl axial bromine in cyclohexyl bromide had yielded cis-&-Butyl0.51 4.81 cyclohexyl 0 44 4.63 the surprisingly small value of 0.7 kcal./mole. cis-4-lIethylIn comparison, the corresponding value for the cyclohexyl G 10 . 49 3.14 3.26 . 3:3 1.21 1.07 2 28 methyl group is generally accepted to be 1.6-1.8 Cyclohexyl kcal./mole,10 despite the fact that methyl and trans-4-8-Butylcyclohexyl 0 , l ~ I ( (0.083)d (0.071)d bromine are usually thought of as being similar in a In 1. m i l k - 1 s'ic: '. r, .lssunied. Infcrred from the size." Infrared studies on cyclohexyl bromide also fact that 1-bromooctanc gave only 2-octyl phenyl thiocther support a low value for E B ~since , fairly intense (in 84% yield) with sodium thiophenolate. These values bands ascribable to the axial as well as the eyua- are low in accuracy. torial form are found in the ~ p e c t r u m . ' ~More~'~ From the data in Table I, the conformational over, the energy difference between the gauche and equilibrium constant K for bromine (cf. Fig. 1, trans form of propyl bromide-which should be equal to 1/2 EBrl'--iS only 0.1-0.5 k ~ a l . / m o l e , ' ~X = Br) can be calculated by means of equation whereas the corresponding value for n-butane is (i) using either substitution rate constants, ks, or elimination rate constant, kE. The value for K 0.S kcal./mole. l 4 from substitution rates is 3.2, that from elimination Results The kinetics of the reaction of cyclohexyl bro- rates 3.6. The two values are in satisfactory agreemide, cis- and trans-4-t-butylcyclohexylbromideI6 ment. b'e shall use an average value of K B r of and cis-4methylcyclohexyl bromideL6with sodium 3.4. This is also the value obtained from the overthiophenolate in 8iY0 ethanol was followed essen- all rate constants ( k T ) which are apt to be more accurate than the apportioned constants k s and tially by the method previously described,s ;.e., kE. The corresponding free energy difference (7) It is assumed t h a t t h e #-butyl group exerts no polar or steric efA F B r is - 0.73 kcal./mole. fect across t h e ring. Evidence f o r t h e correctness of this assumption I t is rather surprising that this free energy difhas been provided 3 . 4 ference is so much smaller than the corresponding ( 8 ) E. L. Eliel and R. S.R o T H I S J O U R N A L , 79, 5995 (1057). (9) E. J . Corey, i b i d . , 76, 2301 (1953). difference for methyl although, as mentioned above, (10) W. G. Dauben a n d K . S. Pitzer in M .Newman's "Steric E f there are a number of observations in the literature fects in Organic Chemistry," J o h n Wiley and Sons, Inc., New York, in agreement with this finding. The same point N. Y.,1950, chapter 1. C
(11) L. Pauling, "The Nnturc of the Chemical Bond," Cornell 1948, p. 189, gives t h e van der Waals University Press, Ithaca, N. P., radius of bromine as 1.95 k . and t h a t of t h e methyl group a s 2.0 A. (12) M . Larnaudie, Comgt rend., 235, 154 (1952); P. Klaeboe, J , J. Lothe and K . Lunde, Acta Chem. Scand., 10, 1465 (1956); S. Mizushima, private communication. A contrary indication comes from molar Kerr constants-R. J. W. LeFkvre and C. G. LeFkvre, Chemistry & Induslry, 55 (1956)-according t o which cyclohexyl bromide appears t o exist almost entirely in t h e equatorial conformation. (13) P . Klaeboe, J. J. Lothe and K. Lunde. Actn Chem. Scund., 11, 1677 (1927), have calculated, by a n admittedly crude method, Ne,". = 0.85/0.lS corresponding to a value of Eer of 1.0 kcal./mole. (14) Reference 10, p 18. (15) C . R o m a k i , I. Ichisima. I(.Kuratani, T , hliyazawa, T. Shimanouchi and S. Ili?ri.;hirna. Bull. C/I,,II?. S o r . Jofinn, 28, ( : C ~ L I ~ ~ G Y L:in-ai. Acad. Sci. U . S . , 44, 998 (1958). (L) F. I.ynen, H. Eggerer, U. Henning and I. Leisel, A i i g c r u . C/zci)t.. 7 0 , 7:18 (1038).
could be distinguished from IsPP by virtue of its extreme acid lability, paper electrophoresis under alkaline conditions failed to separate the product from IsPP. Studies with IsPPa2 indicated that pyrophosphate was released on acidification of the product. It was anticipated that acid cleavage would occur a t the -C-0- linkage and that if the original product were DmalPP, i t would to some extent undergo allylic rearrangement5 to form dimethylvinylcarbinol in addition to dimethylallyl alcohol. Cleavage with phosphatase, which acts a t the -0-P- linkage should yield only dimethylallyl alcohol. Products of both such hydrolyses were analyzed and the results were in accord with that expected from the cleavage of DnialPP (Table 1). Preliminary data indicate that DmalPP is iiicorporated into FaPP and that IsPP-isomerase is the only iodoacetamide-sensitive4 step in thr enzymatic sequence between mevalonic acid and squalene. -1special role is suggested for the sulfhydryl group in the isomerization such as the ( 5 ) R.I I . De\Volle and W. G. Young, Cliem. Reus., 66, 753 (19.58).
